{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00163694","sets":["6164:6165:6242:8743"]},"path":["8743"],"owner":"11","recid":"163694","title":["密結合並列演算加速機構TCAによるGPU対応GASNetの実装と評価"],"pubdate":{"attribute_name":"公開日","attribute_value":"2016-05-30"},"_buckets":{"deposit":"95d60592-50b3-4de2-b631-eca9b180b0a1"},"_deposit":{"id":"163694","pid":{"type":"depid","value":"163694","revision_id":0},"owners":[11],"status":"published","created_by":11},"item_title":"密結合並列演算加速機構TCAによるGPU対応GASNetの実装と評価","author_link":["321285","321283","321280","321281","321277","321276","321282","321278","321284","321275","321286","321279"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"密結合並列演算加速機構TCAによるGPU対応GASNetの実装と評価"},{"subitem_title":"Implementation and Evaluation of GPU-aware GASNet by Tightly Coupled Accelerators","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"通信","subitem_subject_scheme":"Other"}]},"item_type_id":"18","publish_date":"2016-05-30","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_18_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"筑波大学大学院システム情報工学研究科"},{"subitem_text_value":"筑波大学大学院システム情報工学研究科／現在，筑波大学計算科学研究センター"},{"subitem_text_value":"東京大学情報基盤センター"},{"subitem_text_value":"筑波大学計算科学研究センター／現在，日本原子力研究開発機構"},{"subitem_text_value":"筑波大学計算科学研究センター／筑波大学大学院システム情報工学研究科"},{"subitem_text_value":"Lawrence Berkeley National Laboratory"}]},"item_18_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of System and Information Engineering, University of Tsukuba","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of System and Information Engineering, University of Tsukuba / Presently with Center for Computational Sciences, University of Tsukuba","subitem_text_language":"en"},{"subitem_text_value":"Information Technology Center, The University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"Center for Computational Sciences, University of Tsukuba / Presently with Japan Atomic Energy Agency","subitem_text_language":"en"},{"subitem_text_value":"Center for Computational Sciences, University of Tsukuba / Graduate School of System and Information Engineering, University of Tsukuba","subitem_text_language":"en"},{"subitem_text_value":"Lawrence Berkeley National Laboratory","subitem_text_language":"en"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/163694/files/IPSJ-HPCS2016025.pdf","label":"IPSJ-HPCS2016025.pdf"},"date":[{"dateType":"Available","dateValue":"2018-05-30"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HPCS2016025.pdf","filesize":[{"value":"3.0 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"330","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"330","billingrole":"14"},{"tax":["include_tax"],"price":"330","billingrole":"44"}],"accessrole":"open_date","version_id":"9d5b17c6-320d-4839-bea8-07f5bcb3d2cd","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2016 by the Information Processing Society of Japan"}]},"item_18_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"佐藤, 賢太"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"藤田, 典久"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"塙, 敏博"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"松本, 和也"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"朴, 泰祐"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Khaled, Ibrahim"}],"nameIdentifiers":[{}]}]},"item_18_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Kenta, Sato","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Norihisa, Fujita","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Toshihiro, Hanawa","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kazuya, Matsumoto","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Taisuke, Boku","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Khaled, Ibrahim","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_5794","resourcetype":"conference paper"}]},"item_18_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年，GPU のような演算加速装置を用いたクラスタが HPC 分野で多く用いられるようになってきている．筑波大学計算科学研究センターでは，ノードを跨ぐ演算加速装置間での直接通信を実現するために，密結合並列演算加速機構 TCA（Tightly Coupled Accelerators） を提唱している．この TCA の実装として PEACH2（PCI Express Adaptive Communication Hub version 2） が開発されており，ノードを跨ぐ GPU 間での直接通信を行うことができる．しかしながら，TCA/PEACH2 を利用するためには独自の API を用いる必要があり，プログラミングコストが高く，既存のアプリケーションの移植も容易ではないという問題がある．本研究では，PGAS 言語を対象とした通信ライブラリである GASNet に注目し，これを GPU を対象とした PEACH2 に実装する．これによって，GASNet を介して各種のソフトウェアとの互換性が生じ，TCA/PEACH2 が広く利用できるようになると考えられる．既存の GASNet では GPU メモリを対象とした通信しか想定されておらず，GPU を対象とした拡張は現在開発段階にあり，本論文では現在進行中の GASNet の GPU 対応の拡張についても触れる．TCA/PEACH2 による GASNet のプロトタイプ実装において，ノードを跨ぐ GPU 間の通信性能は TCA/PEACH2 を直接使用した場合の性能と比較して，最小レイテンシの増大は 15%程度に抑えられ，ソフトウェア支援によって最大バンド幅は 1.2 倍の性能向上を達成した．","subitem_description_type":"Other"}]},"item_18_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Recently, PC clusters equipped with GPU as accelerators are widely spread and operated. We have been proposing Tightly Coupled Accelerators (TCA) architecture to realize inter-node direct communication among GPUs, and we developed PCI Express Adaptive Communication Hub version 2 (PEACH2) as a prototype of TCA implementation. However, currently non-standard unique API is required to use TCA/PEACH2, so that the programming cost is expensive and porting of existing applications is not easy. On the other hand, GASNet library developed by Lawrence Berkeley National Laboratory provides low-level communication layer for Partitioned Global Address Space (PGAS) languages such as Unified Parallel C (UPC), Co-Array-Fortran and XcalableMP (XMP), and so on. GASNet assumes only CPU memory as communication target, and extension for GPU-aware GASNet is work in progress now. Beside of GPU-aware GASNet development on commodity network such as InfiniBand, we implement it on TCA/PEACH2 to provide general programming and system software porting on this hardware in this paper. We also mention currently planned features of GPU-aware GASNet. In the case of inter-node GPU communication using GASNet prototype implementation by TCA/PEACH2, the minimum latency is increased only 15 % from the case with native API, and the maximum bandwidth is increased by 1.2 times of native API thanks to software support.","subitem_description_type":"Other"}]},"item_18_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"76","bibliographic_titles":[{"bibliographic_title":"ハイパフォーマンスコンピューティングと計算科学シンポジウム論文集"}],"bibliographicPageStart":"68","bibliographicIssueDates":{"bibliographicIssueDate":"2016-05-30","bibliographicIssueDateType":"Issued"},"bibliographicVolumeNumber":"2016"}]},"relation_version_is_last":true,"weko_creator_id":"11"},"id":163694,"updated":"2025-01-20T11:13:43.629381+00:00","links":{},"created":"2025-01-19T00:35:36.381839+00:00"}